Relationship between Obesity and Iron Deficiency in Healthy Adolescents

Abstract

The relationship between overweight/obesity (excess of weight [EW]) and iron deficiency (ID) is not well defined.

Objective:

To analyze the relationship between EW and ID in healthy adolescents, assessing the contribution of new diagnostic measures of iron status and erythropoietic activity.

Method:

A cross-sectional study was made of 405 healthy adolescents, 12–16 years of age. A total of 289 were normal weight (NW) and 116 were otherwise healthy EW. Epidemiological, socioeconomic, diet, BMI Z-score, CRP (C-reactive protein), hematological, iron status, and erythropoietic activity parameters were measured. Statistical tests were Student's, analysis of variance (ANOVA), Chi-square, Pearson's correlation, and odds ratio.

Results:

ID prevalence in the EW group was 22.6% vs. 29.5% in the NW group (p: 0.3). Greater body weight was associated with lower reticulocyte hemoglobin content (CHr) (NW: 31.3 ± 1.7 pg vs. OW: 30.2 ± 1.7 pg, p: 0.007) and greater CRP (NW: 0.1 ± 0.2 mg/dL vs. OW: 0.2 ± 0.18 mg/dL, p < 0.001), leukocytes (NW: 6.69 ± 1.57 × 10³/L vs. OW: 7.43 ± 1.63 × 10³/L, p < 0.02), platelets (NW: 265.6 ± 58.9 × 10³/L vs. OW: 291.8 ± 54.4 × 10³/L, p < 0.002), ferritin (NW: 32.1 ± 17.9 ng/mL vs. OW: 42.8 ± 20.3 ng/mL, p: 0.01), serum transferrin receptor (sTfR) (NW: 1.39 ± 0.4 mg/L vs. OW: 1.73 ± 0.45 mg/L, p: 0.008), sTfR-F index (sTfR/log Ferritin) (NW: 1.06 ± 0.7 vs. OW: 1.33 ± 0.85, p: 0.036) and reticulocyte count (NW: 54.2 ± 18 × 10³/L vs. OW: 65.4 ± 24.2 × 10³/L, p: 0.003). A positive correlation was observed between the BMI Z-score and CRP, ferritin, sTfR, leukocytes, platelets, and reticulocyte count, and a negative one between the BMI Z-score and CHr and medium corpuscular volume.

Conclusions:

The prevalence of ID in otherwise healthy EW adolescents was no higher than in NW. The effect of obesity on iron status was low. The findings reveal the concomitant low-grade inflammation, and probably the effect of adiposity on erythropoietic activity. Specific cutoff values for ID in adolescents with OW need to be defined.

Introduction

Obesity and iron deficiency (ID) remain prevalent. These continuing public health problems can have significant clinical consequences, and the adolescent population is at particular risk.

In recent years, the prevalence of obesity has increased alarmingly worldwide,¹ in parallel with the associated morbidity,^2,3 while ID is still the most commonly found nutritional deficit in industrialized countries and elsewhere.^4,5 During adolescence, iron requirements increase due to the growth spurt, the greater volume of blood and muscle mass, and menstrual blood losses.^6,7 Among newly proposed diagnostic parameters, the concentrations of serum transferrin receptor (sTfR) and reticulocyte hemoglobin (CHr) have proven to be very useful.^8–12 In ID, the concentration of sTfR increases in plasma¹³ and does not change with infection or inflammation.^9,13–15 CHr is also an early marker of functional ID.^13,16,17 Recently, cutoff values for sTfR and CHr diagnoses of ID in healthy children and adolescents with normal weight (NW) have been published, and an alternative definition of ID, with high discriminatory power, has been proposed, which can be used in clinical practice and, above all, which is indicated in situations that may alter the sensitivity of the iron parameters classically used (infection, inflammation, or chronic disease).¹¹

The inverse association between iron and obesity has been described in both adults^18,19 and children and adolescents,^20–25 but it is not currently sufficiently clarified. Many studies observe an increased risk of ID in line with the BMI. This relationship could be explained by: (1) an increase in iron requirements, due to the greater blood volume required by the increased body weight, provoking a real ID; (2) a decrease in the bioavailability of iron due to its sequestration in the reticuloendothelial system (assuming that obesity is a chronic inflammatory state, which would imply functional ID); and (3) both situations.¹⁸

Current evidence identifies hepcidin, a hormonal peptide secreted by the liver and the adipocytes, as a major factor in the alteration of iron metabolism that is observed in subjects with obesity.^18,22,26,27 The proinflammatory cytokines that are characteristic of obesity would induce an increase in the production of hepcidin, thus reducing the absorption of iron in the intestine and its release from the macrophages (causing functional ID from decreased iron delivery for erythropoiesis), and an increase in the synthesis of ferritin in the reticuloendothelial cells²⁷

The results of previous studies of the relationship between ID and obesity in adolescents are not homogeneous.^{14,20,21,23,24,28} In general, they are case/control or cross-sectional studies, in many cases, nonpopulation based, but considering a broad range of variables related to ID and obesity,^29,30 and using different criteria to define ID. One of the most important iron parameters is serum ferritin,^12,16,31 an acute-phase reactant that is positively associated with adiposity, which reduces its sensitivity.^31,32 To avoid bias, the standard recommendation is to use C-reactive protein (CRP) or other iron markers such as sTfR, free erythrocytic protoporphyrin, or CHr, which are less affected in inflammatory situations.^11,16

Finally, although inflammation and adiposity are known to modify erythropoietic activity, little attention has been paid to this factor in studies of the relationship between obesity and ID.^27,33

In view of these considerations, our research aim in this article is to analyze the relationship between obesity and ID in healthy adolescents, assessing the contribution of new diagnostic measures of the state of body iron and the degree of erythropoietic activity.

Methods

A cross-sectional study was conducted on a population-based representative sample in the city of Almería (Spain), between 2007 and 2009.

Subjects

The sample size was calculated with respect to 9823 adolescents 12–16 years of age attending one of the 38 public and private secondary schools in Almería. Assuming an obesity/overweight prevalence of 25% and an ID prevalence of 10%, a 95% confidence interval (95% CI) and 4.5% precision, a total of 260 subjects were required. Using a multistage probability sampling method, 24 classes were selected for analysis, one from each of the four secondary grades at six schools. Of the 509 students in these classes invited to participate, 17% refused. Thus, 422 subjects were finally analyzed.

The criteria for exclusion were: (1) subjects with a hematological or systemic disease; (2) those with present or previous infectious or febrile disease or CRP >0.5 mg/dL.¹⁶

Informed written consent was obtained from the students' parents or legal guardians and from the participants themselves before data collection. The study complied with the stipulations of the 1975 Helsinki Declaration in all respects and was approved by the corresponding Research and Ethics Committee, at Torrecárdenas Hospital (Almería).

Survey

As described in a previous study,³⁴ the subjects' parents or guardians were asked to complete a questionnaire regarding their origin (Spanish or immigrant), socioeconomic level (Class 1: high-income; Class 2: middle-income; Class 3: low-income, based on the European Socioeconomic Classification),³⁵ and parental educational level (primary or less; secondary or university education). Information on dietary variables was obtained from the responses to a semiquantitative food frequency questionnaire.³⁶ Low iron bioavailability diet was defined as one with a consumption of meat or fish and vegetables or fruits less than twice a week. The same criteria were applied to girls and boys.³⁷

The interviewers and researchers were physicians who had previously completed an appropriate training and standardization program. Concordance among the researchers was assessed, and a concordance correlation coefficient of between 0.83 and 0.90 was obtained.

Anthropometric Measurements and Pubertal Status

All subjects underwent a complete physical examination and the following data were collected: (1) Pubertal stage (Tanner stage),³⁸ according to which the subjects were classified as prepubertal (Tanner 1) or pubertal (Tanner 2–5). (2) Anthropometry: weight, height were obtained using calibrated equipment and following standard methods, recording the average of two measurements; the BMI was calculated using the formula weight (kg)/height² (m) with the Z-score of the specific BMI standard deviation (SD) appropriate for age and sex, using national reference population data.³⁹ Overweight (OW) and obesity (OB) were defined using the criteria proposed by the International Obesity Task Force.⁴⁰ For the purposes of the present study, both terms are included in the expression “excess of weight” (EW). The subjects were divided into two groups according to body weight: Group A: Control group, with teenagers of NW (data on this group have been published previously)^10,41 and Group B: teenagers with overweight or obesity, otherwise healthy.

Parameters Measured

Venous blood samples were collected from fasting subjects and the following parameters were measured: hemoglobin, red cell indices, CHr and leukocytes, platelets, and reticulocyte counts using an ADVIA-120 counter (Siemens Healthcare Diagnostics, NY, USA); transferrin, serum ferritin, and CRP, by the immunoturbidimetric method using Tina-quant-Transferrin, Tina-quant Ferritin, and the CRPLX Kits from Roche Diagnostics GmbH (Boehringer Mannheim); serum iron, by colorimetric assay using the Fe Kit from Roche Diagnostics GmbH (Boehringer Mannheim); total iron-binding capacity (TIBC), calculated as 1 mg transferrin/dL × 1.27 = 1 μg IBC/dL; transferrin saturation (TS), calculated as serum iron/TIBC × 100; sTfR, by immunoturbidimetric assay using the Quantex sTfR Kit (Biokit SA, Barcelona); and serum erythropoietin, by enzyme immunoassay using the Quantikine IVD-Human Erythropoietin Kit (R&D Systems, Minneapolis). The sTfR-F index was calculated as sTfR/log Ferritin.⁴²

ID was defined using a combination of ferritin and hemoglobin, together with new diagnostic measures whose cutoff values were obtained previously in Group A.^10,11,41 Three categories were established: (1) iron stores depletion defined as serum ferritin <12 ng/mL); (2) iron-deficient erythropoiesis (IDE), defined as sTFr-F index >1.4 for girls and >1.5 for boys or ≥2 parameters altered (ferritin <12 ng/mL, CHr <28.5 pg or sTfR >1.75 mg/L for girls and >1.95 mg/L for boys); (3) iron-deficient anemia (IDA), defined as IDE and hemoglobin <11.5 g/dL for girls and <12 g/dL for boys. Any of these three categories were considered to constitute ID. In cases of isolated microcytosis, thalassemia traits (α or β) were excluded.

Statistical Analysis

All statistical calculations were performed using SPSS 21.0 software for Windows and Epidat 4.0. Data normality was evaluated by the Kolmogorov/Smirnov test. Non-normally distributed variables were log transformed to achieve normality before analysis. Results are expressed as mean ± SD and 95% CI. The qualitative variables are expressed as percentages. Student's t-test was applied to compare the mean, and the analysis of variance (ANOVA) test was used when more than two groups were compared. When the results were significant, differences between groups were identified using the Bonferroni post hoc test. The prevalence of EW and ID was established for the group total. Groups A and B were compared, in terms of the study variables, using the odds ratio (OR) and the corresponding 95% CI. The association between EW and ID was determined by comparing the prevalence of ID and the mean values of the analytic parameters for groups A and B. A correlation analysis (Pearson test) was used to calculate the association between the BMI Z-score and the analytical parameters. Statistical significance was defined as p < 0.05.

Results

A total of 422 adolescents 11–16 years of age were initially selected for analysis. Of these, four were excluded due to CRP values >0.5 mg/dL, another four due to thalassemic trait α or β, and nine due to the loss of relevant results. Thus, a total of 405 healthy adolescents were finally included in the study.

The mean age of the total study group was 13.9 ± 1.3 years (95% CI 13.8–14 years). The overall prevalence of EW was 28.6% (Group B). Of these, 21.2% were overweight and 7.4% were obese. The overall prevalence of ID was 13.3%. Of these subjects, 49 (12.1%) presented ID without anemia, and only 5 (1.2%) had IDA.

Table 1 describes the general characteristics of the subjects, and compares groups A and B. The prevalence of overweight was significantly higher among the male adolescents (32.8% vs. 23.3% of females) and in those 12–13 years of age (33.7%). For the other variables, the two groups presented comparable values. The prevalence of ID was not higher in the adolescents with EW and no differences were observed in the participants' dietary characteristics. Menstruation had started in 154 of the girls. There were no differences in ID prevalence between prepubertal adolescents and those who had started menstruation (13.6% vs. 15.6%). A low iron bioavailability diet was more frequent in the menstruating girls (39.6% vs. 26.3%), but the difference was not significant.

Table 1.

Characteristics of Subjects and Comparison between Groups A and B

Variables	Total group, n (%)	Group A “NW,” n (%)	Group B “EW,” n (%)	OR (95% CI)	p
Gender
Female	176 (43.5)	135 (76.7)	41 (23.3)	1	0.04
Male	229 (56.5)	154 (67.2)	75 (32.8)	1.6 (1.1–2.5)
Age
12–13 Years	166 (41.0)	110 (63.3)	56 (33.7)	1
14–15 Years	187 (46.2)	141 (75.4)	46 (24.6)	0.64 (0.4–1)	0.06
16 Years	52 (12.8)	38 (73.1)	14 (26.9)	0.72 (0.36–1.4)	0.36
Pubertal status
Tanner 1	30 (7.4)	20 (66.7)	10 (33.3)	1
Tanner 2–5	375 (92.6)	267 (71.6)	106 (28.4)	0.79 (0.36–1.7)	0.57
Menstruating girls
No	22 (12.5)	16 (72.7)	6 (27.3)	1	0.64
Yes	154 (87.5)	119 (77.3)	35 (22.7)	0.78 (0.25–2.1)
Origin
Spanish	333 (82.2)	233 (70.0)	100 (30.0)	1	0.18
Immigrant	72 (17.8)	56 (77.8)	16 (22.8)	0.66 (0.36–1.2)
Socioeconomic level
Class 1	56 (13.7)	43 (79.6)	11 (20.4)	1
Class 2	170 (42.0)	117 (70.5)	49 (29.5)	1.64 (0.78–3.4)	0.19
Class 3	179 (44.3)	121 (69.5)	53 (30.5)	1.7 (0.82–3.6)	0.15
Parental education level
No education/primary	233 (57.5)	160 (69.6)	70 (30.4)	1.25 (0.8–1.9)	0.34
Secondary/University	172 (42.5)	114 (74.0)	40 (26.0)	1
Diet poor in iron
No	256 (63.6)	157 (68.6)	72 (31.4)	1	0.21
Yes	147 (36.4)	98 (74.8)	33 (25.2)	0.73 (0.45–1.2)
Iron deficiency
No	363 (86.7)	248 (70.5)	104 (29.5)	1
Yes	54 (13.3)	41 (77.4)	12 (22.6)	0.7 (0.35–1.38)	0.3

95% CI, 95% confidence interval; Class 1, high-income; Class 2, middle-income; Class 3, low-income; EW, excess of weight; NW, normal weight; OR, odds ratio; p, signification level.

Table 2 shows the analytical parameters obtained, according to body weight. Higher CRP values were observed in the individuals with EW. The hematological parameters did not present statistically significant differences, except the leukocytes and platelets, which were higher in the OB adolescents, and CHr, which was lower. Serum iron and TS were not influenced by body weight. The values for ferritin, sTfR, and sTfR-F index increased significantly with body weight. sTfR values were elevated (above cutoff) in 10% of NW, compared with 19% of EW (p < 0.01; OR: 2.13; 95% CI:1.12–3.9). Regarding erythropoietic activity, reticulocytes increased with body weight, as did erythropoietin, although the latter association was not statistically significant.

Table 2.

Comparison of Hematological, Iron, and Erythropoietic Parameters According to Body Weight

Variables	Group A	Group B
Variables	NW (N: 289), mean ± SD (95% CI)	OW (n: 86), mean ± SD (95% CI)	OB (n: 30), mean ± SD (95%CI)	ANOVA p
Leukocytes ( × 10³/L)	6.69 ± 1.57 (6.5–6.9)	7.09 ± 1.9 (6.7–7.5)	7.43 ± 1.63 (6.8–8.0)	0.02
Platelets ( × 10³/L)	265.6 ± 58.9 (258.5–272.7)	288.3 ± 61.8 (274.4–302.1)	291.8 ± 54.4 (271.4–312.1)	0.002
Erythrocytes ( × 10³/L)	4.8 ± 0.35 (4.7–4.8)	4.9 ± 0.35 (4.8–4.9)	4.8 ± 0,31 (4.7–4.9)	0.2
Hemoglobin (g/dL)	13.8 ± 1.1 (13.7–14.0)	13.9 ± 1.1 (13.7–14.2)	13.8 ± 1.1 (13.3–14.2)	0.61
MCV (fL)	80.3 ± 3.8 (79.9–80.8)	79.8 ± 2.9 (79.2–80.6)	79.4 ± 4.8 (77.6–81.2)	0.3
RDW (%)	12.8 ± 0.74 (12.7–12.9)	12.8 ± 0.6 (12.7–13.0)	13.0 ± 0.7 (12.7–13.2)	0.42
CHr (pg)	31.3 ± 1.7 (31.0–31.4)	31.1 ± 1.7 (30.7–31.4)	30.2 ± 1.7 (29.3–31.1)	0.007
Reticulocytes ( × 10³/L)	54.2 ± 18.0 (51.8–56.2)	59.0 ± 20.4 (54.8–64.4)	65.4 ± 24.2 (56.4–74.5)	0.003
sEPO (mU/L)	7.9 ± 6.8 (7.0–8.7)	8.6 ± 7.4 (6.8–10.3)	10.7 ± 9.9 (7.0–14.4)	0.12
Serum iron (mcg/dL)	106 ± 45.9 (99.4–110)	107.8 ± 38.8 (97.9–114.4)	99.2 ± 34.7 (86.2–112.2)	0.65
TS (%)	35.1 ± 13.8 (33.6–37.0)	35.6 ± 13 (34.0–39.7)	31.5 ± 11.4 (27.2–35.7)	0.33
sFerritin (ng/mL)	32.1 ± 17.9 (29.3–33.8)	37.8 ± 20.7 (34.7–44.6)	42.8 ± 20.3 (36.8–46.9)	0.01
sTfR (mg/L)	1.39 ± 0.4 (1.36–1.46)	1.44 ± 0.4 (1.37–1.55)	1.73 ± 0.45 (1.56–1.9)	0.008
sTfR-Findex	1.06 ± 0.7 (1.0–1.17)	0.97 ± 0.38 (0.9–1.08)	1.33 ± 0.85 (1.01–1.64)	0.036
CRP (mg/dL)	0.1 ± 0.2 (0.06–0.12)	0.21 ± 0.29 (0.08–0.3)	0.21 ± 0.18 (0.14–0.3)	0.001

ANOVA: post hoc (Bonferroni test): Leukocytes: NW/OB: p < 0.05; Platelets: NW/OW: p < 0.008 and NW/OB: p < 0.05; CHr: NW/OB: p < 0.005; Reticulocytes: NW/OB: p < 0.006; sFerritin: NW/OW and NW/OB: p < 0.008; sTfR: NW/OB: p < 0.000; OW/OB: p < 0.002; sTfR-Findex:OW/OB: p < 0.003; CPR: NW/OW and NW/OB: p < 0.003.

Bold values indicate statistical significance.

ANOVA, analysis of variance; CHr, reticulocyte hemoglobin content; CRP, C reactive protein; MCV, mean corpuscular volume; p, signification level; RDW, red distribution wide; sEPO, serum erythropoietin; sTfR, serum transferrin receptor; sTfR, sTfR-F Index: sTfR/log ferritin; TS, transferrin saturation; OW, overweight; OB, obesity; SD, standard deviation.

Table 3 shows the association recorded between the BMI z-score and the analytic parameters studied. There was a significant inverse association between the BMI z-score, medium corpuscular volume (MCV) and CHr, and a positive one between CRP, ferritin, sTfR, and the leukocyte, platelet, and reticulocyte counts.

Table 3.

Association between BMI Z-Score and Hematological and Biochemical Parameters in Healthy Adolescents

Variables	R	p
Leukocytes ( × 10³/L)	0.144	0.006
Platelets (10³/L)	0.17	0.001
Erythrocytes ( × 10⁶/L)	0.07	0.14
Hemoglobin (g/dL)	0.03	0.52
MCV (fL)	−0.11	0.03
RDW (%)	0.09	0.074
CHr (pg)	−0.14	0.006
Reticulocytes ( × 10³/L)	0.12	0.02
sEPO (mU/L)	0.07	0.18
Serum iron (μg/dL)	−0.008	0.88
TS (%)	−0.04	0.43
sFerritin (ng/mL)	0.14	0.005
sTfR (mg/L)	0.17	0.001
sTfR-F Index	0.04	0.41
CRP (mg/dL)	0.43	0.0001

CHr, reticulocyte hemoglobin content; CRP, C reactive protein; MCV, mean corpuscular volume; p, signification level; r, Pearson correlation; RDW, red distribution wide; sEPO, serum erythropoietin; sTfR, serum transferrin receptor; sTfR-F Index, sTfR/log ferritin; TS, transferrin saturation; CHr, sEPO, sFerritin, sTfR, sTfR-F Index and CRP were log transformed before analysis.

Discussion

This nutritional population study was conducted in Almería, a city in SE Spain with a broad representation of social classes and ethnic groups. The analysis focused on 405 healthy adolescents 12–16 years of age, seeking to determine the relationship between EW and ID. Assuming the situation of low-grade inflammation previously described in obese subjects,²⁷ the following measures were implemented to reduce the impact of bias on the results: (1) elevated CRP was taken as exclusion criteria; (2) CHr, sTfR, and sTfR-F index were used to define ID, extrapolating to Group B the diagnostic cutoff values previously published for Group A¹¹; and (3) measures related to erythropoietic activity (EPO, reticulocyte count) were also evaluated. The epidemiological, socioeconomic, and dietary variables presented a similar distribution in both groups, except for age and gender (Table 1). The prevalence of EW was higher among the males and in the group 12–13 years of age.

Our findings did not reveal a higher prevalence of ID in the subjects with EW (Table 1), which corroborates previous research.^14,43,44 However, numerous studies have confirmed the relationship between obesity and ID in children and adolescents^{20–23,25,28,45} and a recent review by Hutchinson²⁹ concluded that there is an increased risk or prevalence of ID with EW, which could be explained by the inflammatory state of excessive adiposity leading to functional ID, rather than by differences in iron intake or bioavailability, as has been reported previously.^22,46–48 In this study, we also found no significant differences in dietary variables between the NW and EW groups. The heterogeneity observed in the diagnostic criteria for ID applied undoubtedly affected the results. Zhao et al.,¹⁹ in a recent meta-analysis, reported that studies in which ferritin was included as a diagnostic criterion for ID found no association between EW and ID, and Huang et al.²⁸ concluded that the relationship between iron status and obesity depended on the indicator used for the diagnosis of ID. In addition, in our study, the exclusion of individuals with high levels of CRP could have selected EW subjects with a less inflammatory component. Nevertheless, CRP values were higher in the EW group, resulting in a heightened inflammatory state. Undetermined hepcidin or inflammatory cytokines could have provided additional information to identify subjects in whom an inflammatory phenomenon, in conjunction with obesity, might have aggravated the iron status.

However, although a higher prevalence of ID in EW adolescents was not observed, there were differences in some hematological and biochemical markers, in relation to body weight (Table 2). In most cases, the differences were between the NW and OB groups. These findings mean that the parameters related to iron metabolism, erythropoietic activity, and inflammation are more severely affected among persons with excessive body weight.

The hemoglobin and erythrocyte indices did not vary with body weight. This finding corroborates previous research^{43,46,47,49,50} and may reflect the fact that an obesity-related inflammatory state and its effect on iron status do not provoke a decrease in the synthesis of hemoglobin.

CHr presented lower values in adolescents with OB than in those with NW/OW. Although this difference was not clinically significant, it might mean that these persons are in a more precarious situation for functional iron than those in states of excess adiposity. The diagnostic value of this parameter in ID is confirmed, and it is not altered by inflammation.^10,11,16 We found no studies with which these results could be compared.

A significant, progressive increase was observed in leukocyte and platelet values in relation to body weight, which is in accordance with previous reports,^43,50 and can be explained by the increased presence of adipocytes in bone marrow, which secrete inflammatory cytokines that cause changes in hematopoiesis by synthesizing the stem cell factor, with the consequent stimulation of myelopoiesis and erythropoiesis by a non-EPO-mediated mechanism.^33,51 Similarly, the reticulocytes presented progressively higher values, with significant differences between NW and OB. There was a progressive increase in EPO in relation to weight, although the difference was not significant, probably due to the variability of the sample. This change could respond to the increase in blood volume in relation to body weight and could also explain the increase observed in reticulocytes. In contrast, in the pathophysiology of anemia of chronic inflammation, a suppression of erythropoietic activity has been described, which is also mediated by cytokines (interleukin [IL]-1 and tumor necrosis factor [TNF]) that interfere in the kidney with the production or activity of EPO.²⁷

Serum ferritin values rose progressively with body weight, as has been reported previously.^14,23,44,48 These findings highlight the role of ferritin as an acute-phase reactant in relation to chronic inflammatory obesity, which reduces its value as a diagnostic marker of ID in obese patients. Some studies have proposed increasing the diagnostic cutoff value⁷ or using correction factors when CRP >0.5 mg/dL.¹⁴

The present study revealed no differences in serum iron or TS levels between persons with NW and OW/OB, which is in line with previous findings.^14,43,44 A decrease in serum iron levels in obese adolescents was first described by Wenzel et al.⁵² and subsequently confirmed in other studies, both with adults^18,19 and with children and adolescents.^21,23,45,48 The inflammatory condition associated with obesity may result in the expression of inflammatory cytokines such as IL-6, IL-1, TNFα, or interferon gamma, which increase the hepcidin produced in the liver and adipocytes, leading to hypoferremia.²⁷ However, some studies have not observed any increase in hepcidin and hence hypoferremia, in obese patients.^43,44

Few studies have analyzed the sTfR levels to determine the relationship between obesity and iron status. sTfR is a marker of functional iron and of the degree of erythropoietic activity, with proven diagnostic value for persons with ID. It is recommended in inflammatory situations, such as obesity because it is not affected by inflammation.^13,15,16,42 Yanoff et al.,¹⁸ in a study of adults, and other studies focusing on children and adolescents,^21,22,25,53 have observed increased sTfR values in obese persons, associating this response to the ID provoked by obesity. However, Ferrari et al.,¹⁴ in a study of European adolescents, found no such differences and concluded that obesity did not impair iron status. In our case, the sTfR values increased with body weight. This trend, together with the decrease in CHr, might confirm the more precarious situation of tissue iron in obesity, but should probably be considered secondary to the increase in erythropoietic activity evidenced by the increase in reticulocytes, as a local consequence of bone marrow adiposity.^33,51 The sTfR-F index, which is of proven diagnostic utility in ID, was also higher in the EW group, although we believe that when obtained from ferritin it may lose sensitivity in persons with EW. To our knowledge, no previous studies have used this marker in obesity, and so there is no reference for comparison.

Finally, the association between the BMI Z-score and the parameters studied was weak in every case (Table 3). However, the inverse association observed with CHr and MCV may mean that EW has a negative impact on the iron status. Moreover, the association with CRP and ferritin would highlight the inflammatory status in persons with obesity, and the association with leukocytes and platelets, sTfR, and reticulocytes might reflect the local hematopoietic effect of obesity.

This study presents some limitations. First, there is no gold standard with which to define ID. Although iron in bone marrow is considered to be the best indicator, this too has its limitations. Moreover, it was not applicable in our study. However, we do propose new measures that are of proven usefulness and are less influenced by inflammation. Second, the presence of hepcidin was not investigated; this could have provided valuable information about which individuals with EW were subject to chronic inflammation that could affect iron metabolism and make them more susceptible to real ID or to poor use of body iron. Third, the food frequency questionnaire was semiquantitative, and so the exact iron intake was not known.

In summary, using new parameters of diagnostic utility, the prevalence of ID in otherwise healthy EW adolescents was no higher than in NW adolescents. In general, the effect of obesity on iron status was low. The study findings reveal the concomitant presence of low-grade inflammation, with the increase in ferritin losing its diagnostic value in ID, and probably the effect of adiposity on erythropoietic activity. In any case, specific cutoff values for ID in adolescents with obesity need to be defined.

Footnotes

Acknowledgment

The authors want to thank all the children and parents involved in the research for their disinterested collaboration.

Authors' Contributions

M.O.-P. contributed by analyzing the data, reviewing the bibliography, and writing the article. M.Á.V.-L. contributed to conception and design, analysis and interpretation of the data, and writing of the article; M.I.A., R.G.-M., and F.L.-M. contributed to the acquisition of data; A.B.-P. and M.M.G critically revised the article and contributed with their final suggestions. All the authors approved the final article for publication.

Funding Information

Funding has not been received.

Author Disclosure Statement

No competing financial interests exist.

References

World Health Organization. Taking action on childhood obesity [internet] Geneva: World Health Organization 2018. www.who.int/iris/handle/10665/274792 (Last accessed June 20, 2020).

Dalmau-Serra

, Alonso-Franch

, Gómez-López

, et al. Obesidad infantil. Recomendaciones del comité de nutrición de la Asociación Española de Pediatría. Parte II. Diagnóstico. Comorbilidades. Tratamiento. An Pediatr (Barc), 2007; 66:294–304.

Galera-Martínez

, Garcia-Garcia

, Vázquez-López

, et al. Prevalence of metabolic Syndrome among adolescents in a city in the Mediterranean área: Comparison of two definitions. Nutr Hosp, 2015; 32:627–633.

Castro-de-Andrade-Cairo

, Rodrigues-Silva

, Carneiro-Bustani

, et al. Iron deficiency anemia in adolescents; a literature review. Nutr Hosp, 2014; 29:1240–1249.

World Health Organization. Worldwide prevalence on anaemia 1993–2005. https://www.who.int/vmnis/anaemia/prevalence/en (Last accessed December 15, 2019).

Ballabriga

, Carrascosa

. Nutrición en la adolescencia. In: Ballabriga A, Carrascosa A (eds), Nutrición en la infancia y adolescencia. Madrid: Ergón, 2006, pp. 525–573.

World Health Organization. Iron deficiency anaemia assessment, prevention and control. A guide for programme managers. 2001. http://www.who.int/.../en/ida_assessment_prevention_control.pdf (Last accessed June 20, 2020).

Vázquez

, Carracedo

, Lendinez

, et al. The usefulness of serum transferrin receptor for discriminating iron deficiency without anemia in children. Haematologica, 2006; 91:134–135.

Infusino

, Braga

, Dolci

, Panteghini

. Soluble transferrin receptor (sTfR) and sTfR/log Ferritin index for the diagnosis of iron-deficiency anemia: A meta-analysis. Am J Clin Pathol, 2012; 138:642–649.

10.

Ibáñez-Alcalde

, Vázquez-López

MÁ

, Ruíz-Sánchez

, et al. Reference values of reticulocyte hemoglobin content in healthy adolescents. J Pediatr Hematol Oncol, 2018; 40:298–303.

11.

Vázquez López

, López-Ruzafa

, Ibáñez-Alcalde

, et al. The usefulness of reticulocyte haemoglobin content, serum transferrin receptor and the sTfR-ferritin index to identify iron deficiency in healthy children aged 1–16 years. Eur J Pediatr, 2019; 178:41–49.

12.

Mattiello

, Schmugge

, Hengartner

, et al. Diagnosis and management of iron deficiency in children with or without aaemia: consensus recommendations of the SPOG Pediatric Hematology Group. Eur J Pediatr, 2020; 179:527–545.

13.

Beguin

Soluble transferin receptor for the evaluation of erythropoiesis and iron status. Clin Chim Acta, 2003; 329:9–22.

14.

Ferrari

, Cuenca-García

, Valtueña

, et al. HELENA Study Group. Inflammation profile in over- weight/obese adolescents in Europe: An analysis in relation to iron status. Eur J Clin Nutr, 2015; 69:247–255.

15.

Vazquez

, Lendinez

, Leyva

, et al. Serum transferrin receptor in children: Usefulness for determinating the nature of anemia in infection. J Pediatr Hematol Oncol, 2006; 28:809–815.

16.

Thomas

, Hinchliffe

, Briggs

, et al. British Committee for Standards Guideline for the laboratory diagnosis of functional iron deficiency. Br J Haematol, 2013; 161:639–648.

17.

Uijterschout

, Domellöf

, Vloemans

, et al. The value of Ret-Hb and sTfR in the diagnosis of iron depletion in healthy, young children. Eur J Clin Nutr, 2014; 68:882–886.

18.

Yanoff

, Menzie

, Denkinger

, et al. Inflammation and iron deficiency in the hypoferremia of obesity. Int J Obesity, 2007; 31:1712–1719.

19.

Zhao

, Zhang

, Shen

, et al. Obesity and iron deficiency: A quantitative meta-analysis. Obes Rev, 2015; 16:1081–1093.

20.

Nead

, Halterman

, Kaezorowski

, et al. Overweight children and adolescents: A risk group for iron deficiency. Pediatrics, 2004; 114:104–108.

21.

Tussing-Humphreys

, Liang

, Nemeth

, et al. Excess adiposity, inflammation, and iron-deficiency in female adolescents. J Am Diet Assoc, 2009; 109:297–302.

22.

Aeberli

, Hurrell

, Zimmermann

. Overweight children have higher circulating hepcidin concentrations and lower iron status but have dietary iron intakes and bioavailability comparablewith normal weight children. Int J Obes (Lond), 2009; 33:1111–1117.

23.

Manios

, Moschonis

, Chrousos

, et al. The double burden of obesity and iron deficiency on children and adolescents in Greece: The Healthy Growth Study. J Hum Nutr Diet, 2013; 26:470–478.

24.

Frelut

, Girardet

, Bocquet

, et al. Impact of obesity on biomarkers of iron and vitamin D status in children and adolescents: The risk of misinterpretation. Arch Pédiatr, 2018; 25:3–5.

25.

Siyaram

, Bhatia

, Dayal

. Hypoferremic state in overweight and obese children. Indian Pediatr, 2018; 55:72–73.

26.

Nemeth

, Ganz

. Regulation of iron metabolism by hepcidin. Annu Rev Nutr, 2006; 26:323–342.

27.

Weiss

, Ganz

, Goodnough

. Anemia of inflammation. Blood, 2019; 133:40–50.

28.

Huang

, Tok

, Lu

, et al. Relationship between being overweight and iron deficiency in adolescents. Pediatr Neonatol, 2015; 56:386–392.

29.

Hutchinson

A review of iron studies in overweight and obese children and adolescents: A double burden in the Young. Eur J Nutr, 2016; 55:2179–2197.

30.

Johnson

, Showell

, Flessa

, et al. Do neighborhoods matter? A systematic review of modificable risk factors for obesity among low socioeconomic status black and Hispanic children. Chilhood Obesity, 2019; 15:71–86.

31.

Khan

, Khan

, Maimoona

, et al. Ferritin is a marker of inflammation rather than iron deficiency in overweight and obese people. J Obes, 2016; 2016:1937320.

32.

Dignass

, Farrag

, Stein

Limitations of serum ferritin in diagnosing iron deficiency in inflammatory conditions. Int J Chronic Dis, 2018:9394060. DOI: 10.1155/2018/9394060

33.

Veldhuis-Vlug

, Rosen

. Clinical implications of bone marrow adiposity. J Intern Med, 2018; 283:121–139.

34.

García-García

, Vázquez

, Rodriguez

, et al. Iodine intake and prevalence of thyroid autoimmunity and autoimmune thyroiditis in children and adolescents aged between 1 and 16 years. Eur J Endocrinol, 2012; 167:387–392.

35.

Harrison

, Rose

. The European Socio-Economic Classification (ESeC) User Guide. Institute for Social and Economic Research, University of Essex: Colcheser, United Kingdom, 2006.

36.

Hernán García

, Cabrera León

, Alaminos Romero

La salud en la población infantil y adolescente en Andalucía. Encuesta andaluza de salud 2003. Marín Vega J, ed. Sevilla: Consejería de Salud; Consejería para la Igualdad y Bienestar Social.

37.

Powers

. Iron requeriments and iron deficiency in adolescents. https://www.uptodate.com/contents/iron-requeriments-and-iron-deficiency-i-adolescents (Last accessed April 20, 2020).

38.

Tanner

JM.

Growth at Adolescence, 2nd ed. Blackwell Scientific Publication: London, Oxford, 1962, Vol. 212, p. 35.

39.

Carrascosa Lezcano

, Fernández García

, Fernández Ramos

, et al. Spanish cross-sectional growth study 2008. Part II. Height, weight and body mass index values from birth to adulthood. An Pediatr (Barc), 2008; 68:552–569.

40.

Cole

, Bellizi

, Flegal

, et al. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ, 2000; 320:1240–1243.

41.

Vázquez-López

, Ibáñez-Alcalde

. Lendínez Molinos F et al. Reference values of serum transferrin receptor and sTfR/log ferritin index in healthy adolescents. J Pediatr Hematol Oncol, 2015; 37:274–280.

42.

Punnonen

, Irjala

, Rajamaki

. Serum transferrin receptor and its ratio to serum ferritin in the diagnosis of iron deficiency. Blood, 1997; 89:1052–1057.

43.

Demircioglu

, Görunmez

, Dagistan

, et al. Serum hepcidinnlevels and iron metabolism in obese children with and without fatty liver: Case-control study. Eur J Pediatr, 2014; 173:947–951.

44.

Chang

, Li

, Lu

, et al. Inteerleukin-10 as a potencial regulator of hepcidin homeostasis in overweight and obese children: A cross-sectional study in Taiwan. Nutrition, 2014; 30:1165–1170.

45.

Pinhas-Hamiel

, Newfield

, Koren

, et al. Greater prevalence of iron deficiency in overweight and obese children and adolescents. Int J Obes, 2003; 27:416–418.

46.

Menzie

, Yanoff

, Denkinger

, et al. Obesity-related hypoferremia is not explained by differences in reported intake of heme and nonheme iron or intake of dietary factors that can affect iron absorption. J Am Diet Assoc, 2008; 108:145–148.

47.

Cepeda-Lopez

, Osendarp

, Melse-Boonstra

, et al. Sharply higher rates of iron deficiency in obese Mexican women and children are predicted by obesity-related inflammation rather than by differences in dietary iron intake. Am J Clin Nutr, 2011; 93:975–983.

48.

Moschonis

, Chrousos

, Lionis

, et al. Healthy Growth Study group. Association of total body and visceral fat mass with iron deficiency in preadolescents: The healthy Growth Study. Br J Nutr, 2012; 108:710–719.

49.

Eftekhari

, Mozaffari-Khosravi

, Shkidfar

. The relationship between BMI and iron status in iron-deficient adolescent Iranian GIRLS. Public Health Nutr, 2009; 12:2377–2381.

50.

Sal

, Yenicesu

, Celik

, et al. Relationship between obesity and iron deficiency anemia: Is there a role of hepcidin?. Hematology, 2018; 23:542–548.

51.

Zhou

, Yu

, Yue

, et al. Bone marrow adipocytes promote the regeneration of stem cells and hematopoiesis by secreting SCF. Nat Cell Biol, 2017; 19:891–903.

52.

Wenzel

, Stults

, Mayer

. Hypoferremia in obese adolescents. Lancet, 1962; 2:327–328.

53.

Nazif

, El-Shaheed

, El-Shamy

, et al. Study of serum Hepcidin as a potential mediator of the disrupted iron metabolism in obese adolescents. Int J Health Sci (Qassim), 2015; 9:172–178.